Mapping With more Data - FEMA

Ricky

11/10/2020

Introduction

In this project, we are going to create our own interactive map to take a look at the public assistance program summary data from FEMA.

the interactive map on FEMA’s website

Load data

We downloaded FEMA data from this link: https://www.fema.gov/openfema-data-page/public-assistance-funded-projects-details-v1

This .csv file include detailed public assistance funded projects information. We selected the the Sever Storm(s), Hurricane and Coastal Storm incident-types in 2009-2018 for this analysis.

Dateset

This dataset is prepared for county-level plot.

DisasterNum	Year	Incident	State	County	ProjectAmount	FederalShareObligated	ToatlObligated
1871	2010	Severe Storm(s)	North Carolina	Madison	5356.79	4017.59	4017.59
1871	2010	Severe Storm(s)	North Carolina	Madison	5163.41	3872.56	3872.56
1871	2010	Severe Storm(s)	North Carolina	Madison	13811.83	10358.87	10358.87
1871	2010	Severe Storm(s)	North Carolina	Madison	3765.00	2823.75	2823.75
1871	2010	Severe Storm(s)	North Carolina	Madison	12318.38	9238.79	9238.79
1871	2010	Severe Storm(s)	North Carolina	Madison	7488.04	5616.03	5616.03

Here are some explanation from FEMA website:

projectAmount(“Project Amount”): The estimated total cost of the Public Assistance grant project in dollars, without administrative costs. This amount is based on the damage survey.

federalShareObligated(“Federal Share Obligated”): The Public Assistance grant funding available to the grantee (State) in dollars, for sub-grantee’s approved Project Worksheets.

totalObligated(“Total Obligated”): The federal share of the Public Assistance grant eligible project amount in dollars, plus grantee (State) and sub-grantee (applicant) administrative costs. The federal share is typically 75% of the total cost of the project.

Mapping

In this plot, we take all years(2009-2018) data, calculated the summary of obligated amount of each county.

# total obligated amount in thousands
hrccJ <- hrcc %>% group_by(Fips) %>% 
  summarise(state = unique(state), county = unique(county), 
            TotalAmount = sum(totalObligated)/1000,
            .groups = "drop") %>%
  rename(GEO_ID = Fips)

Outliers

In order to give an obvious plot of each county, we delete some outlier points.

# outliers
outliers <- hrccJ %>% filter(TotalAmount %in% outValue)
hrccJ <- hrccJ %>% filter(!(TotalAmount %in% outValue))
kable(head(outliers),"html",align = "c")

GEO_ID	state	county	TotalAmount
01003	Alabama	Baldwin	13845.087
01015	Alabama	Calhoun	7403.059
01039	Alabama	Covington	5297.541
01043	Alabama	Cullman	13460.415
01049	Alabama	DeKalb	6700.742
01059	Alabama	Franklin	19455.656

Popups

Also add some popups to show the State, County and Total amount(in thousands).

# popup 
i_popup <- paste0("<strong>State: </strong>", 
                  hrccJoin$state, "<br>", 
                  "<strong>County: </strong>", 
                  hrccJoin$county, "<br>", 
                  "<strong>Total Obligated Amount: </strong>",
                  round(hrccJoin$TotalAmount,2) ,
                  " ","thousands")

Leaflet map

Plot the leaflet.

Inference:

Leaflet css file: Per Liedman

https://gist.github.com/perliedman/9d42f6d1d6583e882df7ae9e95252c0d

Curiosity

Spatial data is everywhere. When we want to compare the subjects from different areas, creating a map is much more intuitive than just looking at the bland data set. And that’s why interactive mapping tools like leaflet is powerful. In the projects we done, we mainly focused on plotting polygons and paths. In leaflet’s official website, we found that there are many others functions that can be done by leaflet. How could we draw the map that best illustrates our data? How could we prevent our map from being too fancy and uninformative?